Abstract
GABAA receptors mediate the action of many clinically important drugs interacting with different binding sites. For some potential binding sites, no interacting drugs have yet been identified. Here, we established a steric hindrance procedure for the identification of drugs acting at the extracellular α1+β3− interface, which is homologous to the benzodiazepine binding site at the α1+γ2− interface. On screening of >100 benzodiazepine site ligands, the anxiolytic pyrazoloquinoline 2-p-methoxyphenylpyrazolo[4,3−c]quinolin-3(5H)-one (CGS 9895) was able to enhance GABA-induced currents at α1β3 receptors from rat. CGS 9895 acts as an antagonist at the benzodiazepine binding site at nanomolar concentrations, but enhances GABA-induced currents via a different site present at α1β3γ2 and α1β3 receptors. By mutating pocket-forming amino acid residues at the α1+ and the β3− side to cysteines, we demonstrated that covalent labeling of these cysteines by the methanethiosulfonate ethylamine reagent MTSEA-biotin was able to inhibit the effect of CGS 9895. The inhibition was not caused by a general inactivation of GABAA receptors, because the GABA-enhancing effect of ROD 188 or the steroid α-tetrahydrodeoxycorticosterone was not influenced by MTSEA-biotin. Other experiments indicated that the CGS 9895 effect was dependent on the α and β subunit types forming the interface. CGS 9895 thus represents the first prototype of drugs mediating benzodiazepine-like modulatory effects via the α+β− interface of GABAA receptors. Since such binding sites are present at αβ, αβγ, and αβδ receptors, such drugs will have a much broader action than benzodiazepines and might become clinical important for the treatment of epilepsy.
Introduction
GABAA receptors are the major inhibitory transmitter receptors in the brain. They are ligand-gated chloride channels composed of five subunits that can belong to different subunit classes. The existence of six α, three β, three γ, one δ, one ε, one π, one θ, and three ρ subunits in the mammalian nervous system generates an enormous diversity of GABAA receptor subtypes with different subunit composition. The majority of GABAA receptors, however, are composed of one γ2, two α, and two β subunits (Olsen and Sieghart, 2008). GABAA receptors are the site of action of a variety of pharmacologically and clinically important drugs, such as benzodiazepines, barbiturates, neuroactive steroids, anesthetics, and convulsants (Sieghart, 1995). Due to the action of these drugs, it is now clear that GABAA receptors are modulating anxiety, the excitability of the brain, muscle tonus, vigilance, learning, and memory.
Binding sites for several of these drugs have already been identified on these receptors (Olsen and Sieghart, 2008), but a recent modeling study (Ernst et al., 2005) indicated the presence of multiple solvent-accessible pockets within GABAA receptors that also could function as possible drug binding sites. Simultaneous drug interaction with several binding sites could explain the extremely complex pharmacology of these receptors. The benzodiazepines are unique among the drugs interacting with GABAA receptors, because they only can enhance or reduce ongoing GABAergic activity in contrast to other drugs that also can directly open the GABAA receptor-associated chloride channel at higher concentrations (Sieghart, 1995). The benzodiazepine binding site is located in the extracellular domain of GABAA receptors, at the α+γ− interface (Sigel, 2002; Ernst et al., 2003), whereas the two GABA binding sites of these receptors are located at the two β+α− interfaces (Smith and Olsen, 1995). So far, the remaining extracellular α+/β− and γ+/β− interfaces have not been systematically investigated as possible drug binding sites. Drugs interacting with the α+β− interface should be able to modulate αβ, αβγ, and αβδ receptors and should thus exhibit a much broader action than the benzodiazepines. Nevertheless, such drugs might also be able to distinguish between different receptor subtypes depending on the exact α and β subunit type forming their binding site.
Here, we established a systematic investigation of putative binding pockets for the presence of possible drug binding sites. By using the substituted cysteine accessibility method for sterically hindering access of a drug to its binding site, we identified the first compound mediating part of its pharmacological effects via the extracellular α1+/β3− interface of GABAA receptors.
Materials and Methods
GABAA receptor subunits and point mutations.
cDNAs of rat GABAA receptor subunits α1, β3, and γ2S were cloned as described previously (Ebert et al., 1996). cDNAs of the rat subunits α2, α3, and α5 were gifts from P. Malherbe (Hoffmann La Roche, Basel, Switzerland). The mutated constructs α1V211C and α1S204C were a gift from E. Sigel (Institute of Biochemistry and Molecular Medicine, Bern, Switzerland) and were generated as described previously (Tan et al., 2007). For the generation of mutated β3 and γ2 subunits, these subunits were subcloned into pCDM8 expression vectors (Invitrogen) as described previously (Tretter et al., 1997). Mutated subunits were constructed by PCR amplification using the wild-type subunit as a template. For this, PCR primers were used to construct point mutations within the subunits by the gene splicing by overlap extension technique (Horton et al., 1993). The PCR primers for β3T60C and β3Q64C contained XmaI and XhoI and primers for β3M115C contained PstI and XhoI or XhoI and XbaI restriction sites, which were used to clone the β3 or γ2 fragments into pCI vectors, respectively (Promega). The mutated subunits were confirmed by sequencing.
Two-electrode voltage clamp.
In vitro transcription of mRNA was based on the cDNA expression vectors encoding for GABAA receptor subunits α1, α2, α3, α5, β1, β2, β3, and γ2 (all from rat) (Ramerstorfer et al., 2010) After linearizing the cDNA vectors with appropriate restriction endonucleases, capped transcripts were produced using the mMESSAGE mMACHINE T7 transcription kit (Ambion). The capped transcripts were polyadenylated using yeast poly(A) polymerase (USB) and were diluted and stored in diethylpyrocarbonate-treated water at −70°C.
The methods for isolating, culturing, injecting, and defolliculating of oocytes were identical with those described by Sigel et al. (1990) and were described previously (Li et al., 2003). Mature female Xenopus laevis (Nasco) were anesthetized in a bath of ice-cold 0.17% Tricain (ethyl-m-aminobenzoat; Sigma) before decapitation and removal of the frog's ovary. Stage 5–6 oocytes with the follicle cell layer around them were singled out of the ovary using a platinum wire loop. Oocytes were stored and incubated at 18°C in modified Barths' medium [88 mm NaCl, 10 mm HEPES-NaOH, pH 7.4, 2.4 mm NaHCO3, 1 mm KCl, 0.82 mm MgSO4, 0.41 mm CaCl2, 0.34 mm Ca(NO3)2] that was supplemented with 100 U/ml penicillin and 100 μg/ml streptomycin. Oocytes with follicle cell layer still around them were injected with an aqueous solution of mRNA. A total of 2.5 ng of mRNA per oocyte was injected. Subunit ratio was 1:1:5 for α1β3γ2 receptors and 1:1 for α1β3 receptors consisting of wild-type or mutated α1 subunit together with wild-type or mutated β3 subunit. After injection of mRNA, oocytes were incubated for at least 24 h for α1β3 receptors and for at least 36 h for α1β3γ2 receptors before the enveloping follicle cell layers were removed. Collagenase-treatment (type IA; Sigma) and mechanically defolliculating of the oocytes were as described previously (Li et al., 2003).
For electrophysiological recordings, oocytes were placed on a nylon grid in a bath of Xenopus Ringer solution (XR; containing 90 mm NaCl, 5 mm HEPES-NaOH, pH 7.4, 1 mm MgCl2, 1 mm KCl, and 1 mm CaCl2). The oocytes were constantly washed by a flow of 6 ml/min XR, which could be switched to XR containing GABA and/or drugs. Drugs were diluted into XR from DMSO solutions, resulting in a final concentration of 0.1% DMSO perfusing the oocytes. Drugs were applied for 30 s before the addition of GABA, which was then coapplied with the drugs until a peak response was observed. Between two applications, oocytes were washed in XR for up to 15 min to ensure full recovery from desensitization. For current measurements, the oocytes were impaled with two microelectrodes (2–3 MΩ) that were filled with 2 m KCl. Maximum currents measured in mRNA-injected oocytes were in the microampere range for all subtypes of GABAA receptors. To test for modulation of GABA-induced currents by drugs, a concentration of GABA that was titrated to trigger 3% of the respective maximum GABA-elicited current of the individual oocyte (EC3), was applied to the cell with various concentrations of drugs. Such low GABA concentrations are widely used in the literature. At GABA EC3, drug effects are larger than at higher GABA concentrations. In addition, GABA EC3 probably is close to the GABA concentration acting on extrasynaptic receptors, which represent a large part of all GABAA receptors in the brain (Farrant and Nusser, 2005). In some experiments, however, the effect of drugs was also investigated at a GABA concentration eliciting 20% of the maximal current (GABA EC20) and, in other experiments, GABA concentration–response curves were performed in the absence or presence of a constant and high drug concentration. MTSEA-biotin solution (2 mm) was freshly made in XR containing the respective GABA-EC3 concentration. Cells were immediately preincubated in MTSEA-biotin solution for 3 min and washed with XR for 5 min. All recordings were performed at room temperature at a holding potential of −60 mV using a Warner OC-725C two-electrode voltage clamp (Warner Instrument) or a CA-1B oocyte clamp or a TEV-200A two-electrode voltage clamp (both from Dagan). Data were digitized, recorded, and measured using a Digidata 1322A data acquisition system (Molecular Devices). Data were analyzed using GraphPad Prism. Data for GABA-dependent dose–response curve were fitted to the equation Y = Bottom + (Top − Bottom)/1 + 10(LogEC50 − X) * nH, where EC50 is the concentration of the compound that increases the amplitude of the GABA-evoked current by 50% and nH is the Hill coefficient. Data are given as mean ± SE from at least three oocytes and two oocyte batches. Statistical significance was calculated using unpaired Student's t test or a one-way ANOVA with Bonferroni's post test.
Results
The pyrazoloquinoline CGS 9895 enhances GABA-induced currents in α1β3 receptors
Both the α+β− interface and the benzodiazepine binding site at the α+γ− interface contain the α+ side (Fig. 1A). A binding site at the α+β− interface might thus also accommodate at least some of the benzodiazepine site ligands. To identify drugs possibly mediating some of their effects via the α+β− interface and to avoid interaction with the benzodiazepine binding site, we used GABAA receptors composed of α1 and β3 subunits only. Such receptors are assumed to be composed of three β and two α subunits (Tretter et al., 1997; Farrar et al., 1999; Baumann et al., 2001) and should thus have two β3+/α1− interfaces (GABA binding sites), two α1+/β3− interfaces, and one β3+/β3− interface, but no benzodiazepine binding site (Fig. 1B). In a screening of >100 benzodiazepine site ligands from different structural classes at 10 μm, the anxiolytic pyrazoloquinoline 2-p-methoxyphenylpyrazolo[4,3−c]quinolin-3(5H)-one (CGS 9895; a gift from Novartis, formerly Ciba-Geigy) (Fig. 2A) (Bennett, 1987) was able to strongly enhance GABA-induced currents in recombinant α1β3 receptors expressed in Xenopus oocytes. Subsequent concentration–response curves indicated that CGS 9895 started to enhance GABA-induced currents (GABA EC3) in these receptors at 1 μm. At 10 μm, this compound was able to stimulate this current up to 400%, and at 30 μm up to 600% (Fig. 2B). CGS 9895, however, did not directly elicit a chloride current in the absence of GABA.
Top view onto the extracellular domain of GABAA receptors. Each subunit features a plus (+) and a minus (−) side. Binding sites for GABA are located at the interfaces formed by the plus side of the β and the minus side of the α subunits. A, α1β3γ2 receptors composed of one γ2, two α1, and two β3 subunits. The binding site for benzodiazepine ligands (Bz) is located at the interface formed by the plus side of the α and the minus side of the γ subunits. The interface of interest (?) is located at the interfaces formed by the plus side of the α and the minus side of the β subunits. B, α1β3 receptors composed of two α1 and three β3 subunits. Instead of a binding site for benzodazepine site ligands, a second interface of interest exists, as well as an interface formed by a β+ and a β− side.
Structure and pharmacology of CGS 9895. A, Structure of CGS 9895. B, Concentration–effect curves of CGS 9895 on α1β3 and α1β3γ2 receptors at GABA EC3 and GABA EC20. At 30 μm, CGS 9895 enhanced GABA-induced currents (at GABA EC3) in α1β3 and α1β3γ2 receptors to 615 ± 61% (n = 5) and 660 ± 49% (n = 5), respectively. These values are not significantly different. At GABA EC20, 30 μm CGS 9895 stimulated currents in α1β3 (n = 4) and α1β3γ2 (n = 6) to 262 ± 5% and 349 ± 22%, respectively. These values are significantly different (p < 0.05). CGS 9895 stimulation of α1β3 and α1β3γ2 receptors at GABA EC3 or EC20 was significantly different at 3 μm (n = 4, p < 0.05), 10 μm (n = 4–10, p < 0.001), and 30 μm (n = 4–6, p < 0.01). C, D, GABA concentration-response curves in the absence (■) or presence (□) of 10 μm CGS 9895. Effects are normalized to the maximum evoked GABA current. CGS 9895 (10 μm) evokes a left shift of the GABA EC50 value from 11 to 4 μm (p < 0.001) at α1β3 receptors (C), and from 73 to 33 μm (p < 0.001) at α1β3γ2 receptors (D). The experiments were performed six to eight times in different oocytes.
CGS 9895 is an antagonist at the benzodiazepine binding site
Since this compound in radioligand displacement studies has been demonstrated previously to exhibit a low nanomolar affinity for the benzodiazepine binding site of GABAA receptors (Yokoyama et al., 1982; Brown and Martin, 1984), we compared its action at α1β3 and α1β3γ2 receptors. CGS 9895 did not modulate GABA-induced currents at nanomolar concentrations in α1β3γ2 receptors but, at micromolar concentrations, elicited a comparable current enhancement in α1β3 and α1β3γ2 receptors (Fig. 2B). This effect could be observed at GABA concentrations eliciting 3% (GABA EC3) or 20% (GABA EC20) of the maximal GABA current in the respective oocytes (Fig. 2B). In addition, 10 μm CGS 9895 elicited a left shift in the GABA concentration–response curve in α1β3 and α1β3γ2 receptors (Fig. 2C,D). Since the CGS 9895 effect was much stronger at GABA EC3 than at GABA EC20, all following experiments were performed at GABA EC3. CGS 9895 thus seems not to need the benzodiazepine binding site for producing its effect on these GABAA receptors. This conclusion was confirmed by experiments indicating that the effect of 10 μm CGS 9895 on α1β3γ2 receptors in contrast to that of 10 μm diazepam could not be inhibited by a 10 μm concentration of the benzodiazepine site antagonist Ro15–1788 (Fig. 3A). It is thus possible that CGS 9895 acts as antagonist at the benzodiazepine binding site and stimulates these receptors via a second, as yet unknown, site, which is also present at α1β3 receptors. This conclusion was confirmed by experiments demonstrating that 50 nm CGS 9895, a concentration that completely saturates the benzodiazepine binding site of GABAA receptors but does not stimulate GABA-induced currents, was able to completely inhibit the effects of diazepam on α1β3γ2 receptors (Fig. 3B). These data are in agreement with previous results indicating that CGS 9895 is an antagonist at the benzodiazepine binding site (Brown and Martin, 1984; Bennett, 1987).
CGS 9895 is an antagonist at the benzodiazepine site and enhances GABA-induced currents via a different site at α1β3γ2 receptors. A, Modulation of GABA EC3 currents of recombinant α1β3γ2 receptors by 10 μm CGS 9895 (top) or 10 μm diazepam (bottom) in the absence or presence of 10 μm Ro 15–1788. The experiments were performed 10 times with comparable results. B, Concentration–response curves for diazepam (X), CGS 9895 (■), and diazepam with 50 nm CGS 9895 (□) on α1β3γ2 receptors. The absence of an effect of a single concentration of CGS 9895 at 50 nm is indicated by ○. conc, Concentration. The experiment was performed four times with comparable results.
A steric hindrance approach indicates that CGS 9895 mediates its action via the α+β− interface
To finally exclude the possibility that CGS 9895 exerts its potentiating action via the α+γ− site in α1β3γ2 receptors, the interaction of this drug with the benzodiazepine binding site was inhibited by using the substituted cysteine accessibility method for introducing a steric hindrance into the pocket. For that, several amino acid residues at the γ2− side of the pocket were selected by using our homology model of GABAA receptors as a guide (Ernst et al., 2003) and individually mutated to cysteines. The effects of these mutations on GABA-induced current and its modulation by diazepam were then investigated in Xenopus oocytes expressing these mutated receptors. Introduction of cysteines into binding pockets in many cases impairs the function of receptors, making interpretation of subsequent steric hindrance experiments difficult. Therefore, we selected those point mutations that produced no or small changes in the properties of the receptors for our experiments. The point mutation γ2M130C, although causing an approximately twofold reduction in GABA potency (Fig. 4A, Table 1), did not significantly change the potency and efficacy of diazepam (Fig. 4C) or CGS 9895 (Fig. 4D) to stimulate the mutated receptor. The apparent reduction in GABA-induced currents indicated in Figure 4B was not significant due to the variability of data, possibly caused by different expression levels of receptors in different oocytes. Wild-type and mutated receptors were then exposed to MTSEA-biotin (2 mm) for 2 min, and GABA-elicited currents as well as their potentiation by diazepam or CGS 9895 were measured before and after MTSEA-biotin exposure (Teissére and Czajkowski, 2001). In wild-type α1β3γ2 receptors, MTSEA-biotin had no significant effects on GABA (Figs. 4A,B), diazepam (Fig. 4C), or CGS 9895 (Fig. 4D). Thus, any effect of this reagent on mutated receptors must have been due to covalent labeling of the introduced cysteines. After MTSEA-biotin treatment, the GABA dose–response curve of the mutated receptor was shifted to the left and the potency of GABA for opening the chloride channel was increased (Fig. 4A,B; Table 1), suggesting covalent labeling of the introduced cysteine and possibly indicating that MTSEA-biotin incorporated into the benzodiazepine site at least partially can cause conformational changes comparable to that of benzodiazepines (Teissére and Czajkowski, 2001). MTSEA-biotin nearly eliminated the stimulation of GABA EC3 by diazepam in the mutated receptor (Fig. 4C), but did not influence the stimulation of this receptor by CGS 9895 (Fig. 4D). Together, these results indicated that CGS 9895, in contrast to diazepam, did not produce its GABA-agonistic effect via the benzodiazepine binding site. In addition, these results demonstrate that the action of a drug at a binding site can be efficiently inhibited by such a steric hindrance approach without influencing the effects of drugs interacting with another binding site.
The effects of a steric hindrance introduced into α1β3γ2 receptors via γ2M130C on GABA, diazepam, and CGS 9895 concentration–response curves. A, GABA effects normalized to the maximum evoked GABA current. B, GABA-evoked currents. Due to the variability of the data (expression levels of receptors), the differences observed in the different concentration–response curves are not significant (n = 3–8). C, Diazepam effects on α1β3γ2 and α1β3γ2M130C receptors in the absence or presence of MTSEA-biotin. Top, Concentration–response curves of diazepam (n = 3–5). MTSEA-biotin caused no significant difference in the diazepam effects on α1β3γ2 receptors, but did cause a significant difference in the diazepam effects on α1β3γ2M130C receptors at 100 nm, 1 μm, and 10 μm (p < 0.05). Bottom, Individual current traces at GABA EC3 in the absence or presence of 10 μm diazepam under the conditions indicated. D, Effects of CGS 9895 on α1β3γ2 and α1β3γ2M130C receptors in the absence or presence of MTSEA-biotin. Top, Concentration–response curves of CGS 9895 (n = 3–8). Bottom, Individual current traces at GABA EC3 in the absence or presence of 10 μm CGS 9895 under the conditions indicated. □, α1β3γ2 receptors; ◊, α1β3γ2 receptors plus MTSEA-biotin; ■, α1β3γ2M130C receptors; X, α1β3γ2M130C receptors labeled with MTSEA-biotin.
GABA EC50 and Hill slopes (nH) of various GABAA receptors in the absence or presence of 2 mm MTSEA-biotin
To investigate whether CGS 9895 is mediating its potentiating action at GABAA receptors via the α+β− interface, we applied a similar steric hindrance approach to this interface. For that, several pocket-forming amino acid residues located at the β3− side were mutated to cysteines. Receptors containing the mutated β3 subunits together with α1 and γ2 subunits were expressed in Xenopus oocytes and investigated for the effect of the mutation on GABA-induced currents and their stimulation by diazepam or CGS 9895. The mutation β3Q64C did not significantly change the potency of GABA for enhancing GABA-induced currents in α1β3Q64Cγ2 receptors (Fig. 5A; Table 1) and also did not cause a significant change in GABA-induced currents (Fig. 5B). The high variability of the currents observed might have been due to differences in the expression levels of receptors in individual oocytes. Whereas this mutation did not significantly reduce the potency and efficacy of diazepam for stimulation of these receptors (Fig. 5C), the effects of CGS 9895 were significantly reduced by the point mutation in α1β3Q64Cγ2 receptors (Fig. 5D). On incubation with MTSEA-biotin, GABA-induced currents (Fig. 5A,B), as well as their stimulation by diazepam (Fig. 5C), were not significantly changed, but the effect of CGS 9895 was further reduced (Fig. 5D), suggesting that CGS 9895 might exert its action via the α+β− interface.
The effects of a steric hindrance introduced into α1β3γ2 receptors via β3Q64C on GABA, diazepam, and CGS 9895 concentration-response curves. A, GABA effects normalized to the maximum evoked GABA current. B, GABA-evoked currents. Each concentration–response curve was performed four or five times in independent experiments. Due to the variability of data, there was no significant difference in the GABA-evoked currents under the conditions indicated. C, Diazepam effects on α1β3γ2 and α1β3Q64Cγ2 receptors, the latter in the absence or presence of MTSEA-biotin. Top, Concentration–response curves of diazepam (n = 4–5). MTSEA-biotin caused no significant difference in the diazepam effects under the three conditions. Bottom, Individual current traces at GABA EC3 in the absence or presence of 10 μm diazepam under the conditions indicated. D, Effects of CGS 9895 on α1β3γ2 and α1β3Q64Cγ2 receptors, the latter in the absence or presence of MTSEA-biotin. Top, Concentration–response curves of CGS 9895 (n = 3–10). There was a significant reduction of the effects of 10 μm CGS 9895 in the point mutated receptor (p < 0.05), which was enhanced in the presence of MTSEA-biotin (p < 0.001). Bottom, Individual current traces at GABA EC3 in the absence or presence of 10 μm CGS 9895 under the conditions indicated. □, α1β3γ2 receptors; ◊, α1β3γ2 receptors plus MTSEA-biotin; ■, α1β3Q64Cγ2 receptors; X, α1β3Q64Cγ2 receptors labeled with MTSEA-biotin.
If this is the case, it should be possible to also block the action of CGS 9895 via the α1 subunit. To avoid an interaction of CGS 9895 with both the α+β− interface and the benzodiazepine binding site located at the α+γ− interface, we repeated the steric hindrance approach with GABAA receptors composed of α1 and β3 subunits only. As shown in Figure 6A, the GABA-stimulatory effect of CGS 9895 in α1β3Q64C was not different from that in α1β3 receptors, but covalent modification of the introduced cysteine by MTSEA-biotin was able to reduce this effect by 50%. In contrast, MTSEA-biotin did not significantly change the effects of CGS 9895 in wild-type α1β3 receptors (Fig. 6A). To identify α1+ amino acid residues suitable for this experiment, we mutated several pocket-forming amino acid residues located at the α1+ side to cysteines. The mutation α1V211C did not significantly change the potency of GABA (data not shown) and did not change the effects of CGS 9895 on GABA-induced currents (Fig. 6B). On covalently labeling the introduced cysteine by MTSEA-biotin, however, the CGS 9895-induced enhancement of GABA currents was reduced by 50% (Fig. 6B). Similarly, modification of position α1S204C by MTSEA-biotin also reduced CGS 9895 stimulation (data not shown). Finally, we investigated the effect of CGS 9895 in α1V211Cβ3Q64C receptors, in which cysteines were introduced at the α1+ as well as at the β3− side. As shown in Figure 6C, the two introduced cysteines did not significantly influence the GABA-stimulatory effect of CGS 9895, but covalent labeling of cysteines with MTSEA-biotin nearly completely abolished this effect. In contrast, the effects of ROD 188 (Thomet et al., 2000), interacting with an as yet unknown binding site at the GABAA receptors (Fig. 6D), or of the steroid α-tetrahydrodeoxycorticosterone (data not shown) on α1β3 receptors, were neither changed by the introduced cysteines at the α1+ or β3− side, nor by their covalent modification by MTSEA-biotin, suggesting that receptors were not nonspecifically inactivated by MTSEA-biotin. Together, these data strongly suggest that CGS 9895 exerts its action via the extracellular part of the α1+β3− interface.
Steric hindrance introduced via β3Q64C, α1V211C, or both mutations inhibits the effect of CGS 9895 but not that of ROD 188 on α1β3 receptors. A, Concentration–response curves of CGS 9895 on α1β3 (○), α1β3 receptors plus MTSEA-biotin (◊), α1β3Q64C (■), and α1β3Q64C receptors labeled by MTSEA-biotin (X). Each experiment was performed 5–11 times in different oocytes. The reduction of the CGS 9895 effect by MTSEA-biotin in α1β3Q64C receptors was significant at 10 μm (p < 0.01). B, Concentration–response curves of CGS 9895 on α1β3 (○), α1β3 receptors plus MTSEA-biotin (◊), α1V211Cβ3 (■), and α1V211Cβ3 receptors labeled by MTSEA-biotin (X). The experiments were performed 5–11 times in different oocytes. The reduction of the CGS 9895 effect by MTSEA-biotin in α1V211Cβ3 receptors was significant at 10 μm (p < 0.001). C, Top, Concentration–response curves of CGS 9895 on α1β3 (○), α1β3 receptors plus MTSEA-biotin (◊), α1V211Cβ3Q64C (■), and α1V211Cβ3Q64C receptors labeled by MTSEA-biotin (X). The experiments were performed 5–10 times in different oocytes. The reduction of the CGS 9895 effect by MTSEA-biotin in α1V211Cβ3Q64C receptors was significant at 10 μm (p < 0.05). Bottom, Individual current traces at GABA EC3 in the absence or presence of 10 μm CGS 9895 under the conditions indicated. D, Top, Concentration–response curves of ROD 188 on α1β3 (○), α1β3Q64C (▴), α1V211Cβ3 (■), α1V211Cβ3Q64C (▾), MTSEA-biotin labeled α1β3Q64C (▵), MTSEA-biotin labeled α1V211Cβ3 (□), and MTSEA-biotin labeled α1V211Cβ3Q64C (▿) receptors. The experiments were performed three to five times in different oocytes and there were no significant differences between the curves (wild type vs mutated receptors, or mutated receptors vs mutated receptors in the presence of MTSEA-biotin). Bottom, Individual current traces at GABA EC3 in the absence or presence of 10 μm ROD 188 under the conditions indicated.
In a different approach, we considered the possibility that the presence of CGS 9895 inhibits covalent labeling of the introduced cysteines by MTSEA-biotin. However, due to the irreversible nature of the covalent modification by MTSEA-biotin, which overrules the reversible binding of CGS 9895 in a very short time, and due to simultaneous and partially opposite changes in currents induced by GABA and CGS 9895 during this reaction, such experiments cannot be definitively interpreted.
To investigate whether the effect of CGS 9895 was mediated via the β3+β3− interface that also is present in α1β3 receptors, Xenopus oocytes were injected with β3 subunits only, and the resulting homo-oligomeric receptors were investigated. In agreement with previous results, the ion channel formed by these homo-oligomeric β3 receptors were open in the absence of GABA and could be modulated by pentobarbital and blocked by picrotoxin (Slany et al., 1995; Wooltorton et al., 1997). In contrast, CGS 9895 was not able to change the currents mediated by this channel (data not shown), indicating that either no binding site for this compound can be formed by the β3+β3− interface in these receptors or that CGS 9895 modulation of this channel via this interface is not possible.
In previous studies (Walters et al., 2000), it was demonstrated that diazepam at concentrations of >10 μm was able to potentiate GABA-induced currents in α1β2 receptors. We demonstrated a similar potentiation of GABA-induced currents by 30–100 μm diazepam on α1β3 receptors. The effects at 100 μm diazepam were similar in α1β3 and α1V211Cβ3Q64C receptors but, in contrast to the CGS 9895 effects, could not be significantly inhibited by MTSEA-biotin (data not shown). This is consistent with the suggestion that this low-potency diazepam effect might be caused by an interaction with the transmembrane domain of the receptor (Walters et al., 2000; Baur et al., 2008).
While this work was in progress, evidence was presented indicating that flurazepam, which has been shown to interact with at least two different binding sites in α1β2γ2 receptors (Walters et al., 2000), not only enhanced GABA-induced current via the benzodiazepine binding site, but at concentrations >10 μm also inhibited this current by interacting with the α1+/β2− interface (Baur et al., 2008). We thus investigated whether flurazepam causes its inhibiting effect at high concentrations via the CGS 9895 binding site at the α1+β3− interface. Results indicated that flurazepam at 250 μm was able to inhibit the GABA-stimulatory effect of 10 μm CGS 9895 in α1β3 receptors from 413 ± 25% to 167 ± 28% (n = 4, p < 0.001). In dose–response experiments, flurazepam exhibited no effects in α1β3 receptors up to 1 μm. At higher concentrations, this compound dose dependently inhibited GABA EC3, indicating that, at this receptor, flurazepam in high concentrations behaves as a negative allosteric modulator (data not shown). This negative modulatory effect of flurazepam, however, could not be inhibited by covalently modifying residue β3Q64C, α1V211C, or both by using MTSEA-biotin (data not shown). In addition, this negative modulatory effect at α1β3 receptors, in contrast to the positive modulatory effect of CGS 9895, could not be inhibited by the covalently labeling of residue α1S204C using MTSEA-biotin (data not shown). The absence of inhibition of the negative allosteric effect of high concentrations of flurazepam by all these covalently modified residues seems to indicate that the negative allosteric effect of flurazepam in α1β3 receptors is not mediated via the extracellular α1+β3− interface.
CGS 9895 differentially modulates different receptor subtypes via its binding site at the α+β− interface
To investigate a possible receptor subtype selectivity of CGS 9895, receptors containing the β3 and different α subunits were investigated. As shown in Figure 7A, CGS 9895 exhibited the highest stimulation of GABA-induced current in receptors composed of α1β3 subunits. In receptors composed of α2β3, α3β3, or α5β3 subunits, the effect of CGS 9895 for stimulating GABA-induced currents was only half of that for α1β3 receptors. Similar results were obtained in receptors composed of αxβ3γ2 subunits. At 10 μm, the effects of CGS 9895 on GABA-induced current in α1β3γ2 receptors was approximately twice as high as that on receptors composed of α2β3γ2, α3β3γ2, or α5β3γ2 subunits (Fig. 7C). Interestingly, however, in contrast to α1β3γ2, α2β3γ2, and α5β3γ2 receptors, which could not be modulated at low CGS 9895 concentrations, receptors composed of α3β3γ2 subunits could be weakly stimulated (up to 130% at 100 nm) at nanomolar concentrations, indicating that CGS 9895 is a weak positive allosteric modulator of α3β3γ2 receptors at these concentrations. Since this effect could be inhibited by covalent labeling of γ2M130C with MTSEA-biotin in α3β3γ2M130C receptors (Fig. 7D), this weak positive modulatory effect was mediated via the benzodiazepine binding site. These data indicate that CGS 9895 is an antagonist at the benzodiazepine binding site of α1β3γ2, α2β3γ2, or α5β3γ2 receptors, but a weak positive allosteric modulator at the benzodiazepine site of receptors composed of α3β3γ2 subunits.
CGS 9895 effects in GABAA receptors containing different α or β subunits. A, Concentration–response curves of CGS 9895 on α1β3 (■), α2β3 (▴), α3β3 (▾) and α5β3 (●) receptors. The experiments were performed 3–11 times in different oocytes. The effects of 10 μm CGS 9895 were significantly different for α1β3 receptors (p < 0.01, one-way ANOVA with Bonferroni's post test). B, Concentration–response curves of CGS 9895 on α1β1 (▿), α1β2 (○), α1β3 (■), and α1β1S290N (▾) receptors. The experiments were performed 4–11 times in different oocytes. The effects of 10 μm CGS 9895 were significantly different for α1β1, but not for the other receptors (p < 0.001). C, Concentration–response curves of CGS 9895 on α1β3γ2 (■), α2β3γ2 (▴), α3β3γ2 (♦), and α5β3γ2 (▾) receptors. The experiments were performed 3–10 times in different oocytes. The effects of CGS 9895 at 10 and 100 nm were significantly different at α3β3γ2 compared with the other receptors (p < 0.05). The effects of 10 μm CGS 9895 were significantly different only between α1β3γ2 and α5β3γ2 receptors (p < 0.01, one-way ANOVA with Bonferroni's post test). D, Concentration–response curves of CGS 9895 on α3β3γ2 (♦), α3β3γ2M130C (■), and MTSEA-biotin labeled α3β3γ2M130C (X) receptors. The effect of 10 and 100 nm CGS 9895 at α3β3γ2 receptors was reduced at α3β3γ2M130C and MTSEA-biotin labeled α3β3γ2M130C receptors. The experiments were performed four to six times in different oocytes.
We also investigated the influence of the type of the β subunit on the CGS 9895 stimulation of GABA-induced current. Results obtained indicated that this compound exhibits a comparable stimulation of GABA-induced currents in α1β3 and α1β2 receptors, but nearly no stimulation in α1β1 receptors (Fig. 7B). The lack of effect of CGS 9895 on β1-containing receptors is similar to that of several other GABA modulators, such as loreclezole, etomidate, and DMCM (Stevenson et al., 1995). Since their β selectivity can be modulated by a point mutation of the amino acid residue 290, which is an Asn in the β2 or β3 subunit and a Ser in the β1 subunit, we investigated whether the mutation β1S290N can enhance the effect of CGS 9895 on α1β1 receptors. As shown in Figure 7B, 10 μm CGS 9895 were able to modulate GABA-induced currents in α1β1 receptors containing the point mutation β1S290N to a similar extent as in α1β2 or α1β3 receptors. Since residue β1S290N is located within TM2 of this subunit and is far away from the putative binding site of CGS 9895 in the extracellular domain (distance of β3N290 to α1V211C or β3Q64C is approximately twice the length of MTSEA-biotin), these data suggest that, in this case, N290 probably is not involved in direct binding of this compound but in the transduction of its effect on GABA-induced chloride current.
Discussion
Steric hindrance: a novel tool for the identification of drug binding sites
A variety of clinically important drugs exert their effects via GABAA receptors, but additional drug binding sites might be present at these receptors (Ernst et al., 2005). To develop a tool for identifying possible drug binding sites at GABAA receptors, we established a steric hindrance approach. In a proof of principle, we introduced a cysteine at the γ2− side (γ2M130C) of the benzodiazepine binding pocket and demonstrated that covalently labeling of this cysteine with MTSEA-biotin in α1β3γ2M130C receptors drastically reduces the ability of diazepam to stimulate GABA-induced ion flux without altering the effects of drugs interacting with other binding sites of these receptors. This was expected because MTSEA-biotin, with its length of 12 Å (Teissére and Czajkowski, 2001), when irreversibly bound to a cysteine located within a pocket, cannot reach into another binding pocket. Due to its length and flexibility, however, this compound cannot be used for exactly localizing the binding site within the pocket.
We then used a similar approach for identifying drugs interacting with the α1+β3− interface by introducing cysteines at the α1+ (α1V211C) and β3− (β3Q64C) sides and subsequently investigating whether covalent labeling of these cysteines by MTSEA-biotin could inhibit the effects of the drugs. The pyrazoloquinoline CGS 9895, which previously has been demonstrated to exhibit a high affinity for the benzodiazepine binding site of GABAA receptors (Yokoyama et al., 1982; Brown and Martin, 1984), was selected for these investigations because this drug was able to enhance GABA-induced currents in receptors containing α1 and β3 subunits only. Several lines of evidence indicated that CGS 9895 exerts its GABA-enhancing action in α1β3γ2 receptors via the α1+β3− interface. First, this drug at nanomolar concentrations does not modulate GABA-induced currents, but completely prevents the action of diazepam in α1β3γ2 receptors, indicating that it acts as an antagonist at the benzodiazepine binding site (Bennett, 1987). Second, only at micromolar concentrations was CGS 9895 able to enhance GABA-induced currents in α1β3γ2 receptors, but this effect could not be inhibited by the benzodiazepine site antagonist Ro15–1788, again indicating that it is not mediated via the benzodiazepine binding site. Third, CGS 9895 was able to elicit a comparable stimulation of GABA-induced currents in receptors composed of α1β3 or α1β3γ2 subunits, but not in homo-oligomeric receptors composed of β3 subunits only. Fourth, the inability of this drug to elicit a current in the absence of GABA and the GABA-enhancing effect of this drug argue against its interaction with the GABA binding sites located at the α1− β3+ interfaces. Fifth, the GABA-enhancing effect of CGS 9895 could be partially inhibited by a steric hindrance mediated by MTSEA-biotin via cysteines introduced at the α1+ (α1V211C) and β3− (β3Q64C) side, and was even more inhibited when the steric hindrance was introduced via both the α1+ and the β3− side, finally supporting the conclusion that the respective binding site is located at the extracellular part of the α1+β3− interface.
It has been shown previously that some benzodiazepine site ligands can also interact with GABAA receptors via binding sites different from the classical benzodiazepine site. For instance, Hauser et al. (1997) demonstrated that [3H]flunitrazepam exhibits a high affinity binding site at α6β2γ2 receptors, although at very high concentrations, this compound was able to displace [3H]Ro15-4513 binding from the respective benzodiazepine binding site. This high-affinity flunitrazepam binding obviously elicited a negative modulatory effect at these receptors. In other experiments, it was demonstrated that the negative allosteric modulator [3H]Ro15-4513 exhibited a high affinity binding to α4/6β3δ GABAA receptors that could be displaced by several other benzodiazepine site ligands (Hanchar et al., 2006). Finally, Walters et al. (2000) reported that, in α1β2γ2 receptors, diazepam at GABA EC3 not only enhanced GABA-induced currents at a nanomolar concentration via the benzodiazepine binding site, but also exhibited an additional stimulation of GABA-induced currents at micromolar concentrations, which, in contrast to the nanomolar component, could not be inhibited by the benzodiazepine antagonist flumazenil and was also present in α1β2 receptors. This site was possibly located within the membrane-spanning region of the receptor because it could be eliminated by introducing three point mutations into this region. Walters et al. (2000) also demonstrated that flurazepam at high micromolar concentrations produced an inhibitory effect at α1β2γ2 receptors and speculated that this might reflect a desensitization of the GABA current. Finally, Baur et al. (2008) provided some evidence that the site mediating the inhibitory action of flurazepam might be located at the extracellular α1+β2− interface of GABAA receptors. Here, we demonstrated that flurazepam could inhibit the GABA-enhancing effect of CGS 9895 in α1β3 receptors. Actually, flurazepam at high micromolar concentrations acted as a negative allosteric modulator at α1β3 receptors, but this effect could not be inhibited by covalently modifying α1V211C, α1S204C, or β3Q64C by MTSEA-biotin, suggesting that this flurazepam effect was mediated via a site different from that of CGS 9895 in α1β3 receptors. This finding and the multiple binding sites possibly present at GABAA receptors (Ernst et al., 2005) underscore the importance of the present approach, providing a tool for localizing individual drug binding sites in GABAA receptors.
The anxiolytic actions of CGS 9895 seem to be mediated via the benzodiazepine binding site of GABAA receptors containing α3 subunits
CGS 9895 has been demonstrated previously to be a potent ligand for the benzodiazepine binding site of GABAA receptors. In vivo, it elicited its anti-anxiety effect at a dose comparable to that of diazepam (Bennett, 1987). The anxiolytic effect of CGS 9895 thus cannot have been elicited via the micromolar effect of this compound at α1β3γ2 receptors. We therefore investigated the effects of CGS 9895 on various other GABAA receptor subtypes. Results indicated that CGS 9895 at nanomolar concentrations exhibited a weak positive allosteric effect on α3β3γ2 receptors that could be completely inhibited by MTSEA-biotin in α3β3γ2M130C receptors, suggesting that it was mediated via the benzodiazepine binding site of these receptors. This effect at nanomolar concentrations could not be observed in α1β3γ2, α2β3γ2, or α5β3γ2 receptors, supporting the assumption that the anxiolytic effect of this compound is mediated via α3β3γ2 receptors. The absence of muscle relaxant and sedative, as well as locomotor effects of CGS 9895 (Brown and Martin, 1984; Bennett, 1987), can be explained by the absence of effects at GABAA receptors containing α1 or α5 subunits at low concentrations. In addition, the observation that CGS 9895 is able to antagonize the anticonvulsant (Brown et al., 1984) or muscle incoordination effects of diazepam in the rotarod test (Bennett, 1987) at low concentrations is consistent with the antagonistic effects of this compound on the action of diazepam on α1β3γ2 receptors. In contrast, CGS 9895 at 100 mg/kg per os caused a 30% protection against the pentylenetetrazole-induced convulsions in the rat (Bennett, 1987). At such concentrations, plasma levels of CGS 9895 between 20 and 50 μm were reached and thus, this effect could also have been generated via the effects of this compound elicited at micromolar concentrations on α1β3γ2 receptors.
Drugs interacting with the α+β− interface are candidates for a novel GABAA receptor pharmacology
Although most of the actions of CGS 9895 at low drug concentration seem to be mediated via the benzodiazepine binding site of GABAA receptors, drugs interacting via the extracellular α1+β3− interface should exhibit highly interesting properties. Judged by the observation that CGS 9895, even at high concentrations, can only enhance GABA-induced currents but not directly activate these receptors, drugs acting via the α+β− interface will only have GABA-modulatory properties, like the benzodiazepines do. In contrast to the benzodiazepines, however, these drugs will interact with receptors composed of αβ, αβγ, and αβδ subunits and should thus exhibit a much broader anticonvulsive action than benzodiazepines. In addition, these drugs will exhibit receptor subtype-selective actions depending on the exact α or β subunit type forming at the binding site and thus dramatically expand the pharmacology of GABAA receptor subtypes and our ability to modulate specific receptor subtypes. This conclusion is supported by the present results indicating a difference in the ability of CGS 9895 for stimulating GABA-induced currents on αxβy receptors containing different α or β subunits. CGS 9895 is the first compound unequivocally identified to modulate GABAA receptors via the extracellular α1+β3− interface. Although its potency for interacting with this site is relatively low, this compound now can be used as a screening tool for the identification of other possible candidates interacting with this binding site. This approach is especially important for the identification of possible antagonists at this binding site. At present, in the absence of a direct effect of such compounds at GABAA receptors, they only can be detected by their inhibition of the effects of CGS 9895. In addition, using the cysteine mutations α1V211C, α1S204C, and β3Q64C, and their modification by MTSEA-biotin, compounds that cause a positive or negative modulation of GABAA receptors via this binding site can now be identified and that might lead to the development of clinically important drugs with a different spectrum of action.
Footnotes
-
This work was supported by project P19653 of the Austrian Science Fund (to M.E.) and by the FP7 project HEALTH-F4-2008-202088 (Neurocypres; to W.S.). We thank E. Sigel for helpful discussions.
- Correspondence should be addressed to Dr. Margot Ernst, Department of Biochemistry and Molecular Biology, Center for Brain Research, Medical University Vienna, Spitalgasse 4, 1090 Vienna, Austria. Margot.Ernst{at}meduniwien.ac.at
